>Even component prefix is not always uniques. See IKJnnnnn messages. Some >of them belongs to RACF.
How do you know they belong to RACF? I admit IKJ messages sometimes appear where you wouldn't expect TSO routines to be involved. I guess that has "historically grown". Anyway, no two different modules should issue the same numbered IKJ (or whatever) message. >It need not to be unique. It is *strongly suggested*. It is >*convenient*. But the system and systems programmers can surviver >without full uniquity. Component prefixes need to be unique, otherwise a module of one component can inadvertantly be replaced by another module from another component, both residing in the same load library. >Mainframe has a lot of such "standards". For example: Snnnnn libraries >are target libraries (DDDEFs), while Annnn libraries are distribution >libraries. *Usually*. With a number of exceptions, like LINKLIB or >LPALIB (no S). Many of them are exceptions because they existed before >those rules arised. These again aren't component prefixes. It's a naming convention, and I agree, there very often are exceptions to conventions. Peter Hunkeler CREDIT SUISSE ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html